Sunday, November 30, 2008

How can brick-and-mortar book stores compete?: Part 1

Last week, Borders and Barnes & Noble declared their third quarter results. Both companies reported a net loss at an operating level. Both companies are shutting some unprofitable book stores and trying to rein in costs to become profitable.

I am a big fan of both Borders and Barnes & Noble, and cannot help thinking: How can these companies use analytics to survive the onslaught of Amazon.com?

I analyzed the Q3 numbers of B&N, Borders and Amazon, and here's what I found:
  • Online sales account for less than 10% of Barnes & Noble sales and less than 2% of Borders' sales. Of course they account for 100% of Amazon.com's sales.
  • Amazon is growing much faster than B&N and Borders. In Q3, these companies had comparable sales growth of 19%, -6% and -12% respectively. Note that Amazon numbers are for Media sales (Books, DVD, music). Amazon Media sales for international markets grew faster at 24% compared to US sales at 15%.
  • Amazon runs an operating profit whereas its competitors run a loss. The key difference is in SG&A cost and depreciation. Surprisingly B&N has a lower COGS (cost of goods sold) than Amazon. Amazon has a higher operating margin in its international operations. Borders has a higher operating loss margin in small-format Walden US stores (-14.4%) and a lower operating loss margin in international stores (-5.3%).
  • A large format book retail store requires sales of $20 per sq.ft per month to turn an operational profit. B&N is currently registering $18 per sq.ft per month and Borders US is selling only $14 per sq ft per month in its large stores. A Walden's (small format) store requires sales of more than $30 per sq. ft per month to have an operational profit. Currently Walden's US has sales of $17 per sq ft per month only. Thus there is an urgent need to increase sales per sq ft
  • Amazon has much lower inventory than B&N and Borders (as shown below). Borders is carrying more than 5 months of inventory whereas B&N is carrying more than 4 months of inventory. I feel that book stores need to better align inventory with demand.
Here are my suggestions (which I will detail in subsequent posts):
  • B&N and Borders need to create a more credible online option to compete against Amazon.
  1. I recommend usage of collaborative filtering to offer instant book recommendations to online buyers. Also bn.com and borders.com can improve in terms of linking past browsing behavior to book recommendations online
  2. Barnes & Noble needs to consider matching amazon.com on pricing (given it's lower Cost of Goods). E.g., I recently found that the Buffett biography : The Snowball is priced at least 5% lower on Amazon.com
  3. Many customers feel that Amazon has a wider selection of technical and children's books titles than other online book stores. Catering to the long tail may be essential to competing with Amazon.
  • Use brick-and-mortar as a competitive advantage:
  1. Offer the option to buy online and pickup at a store within a few hours
  2. Let loyal book buyers work for you by organizing 'low-cost' book discussions/reading sessions especially for children, niche interest segments
  3. Offer bundled offers in the store, based on basket analysis of previous purchases
  4. Consider installing digital displays in the store showing reader reviews and suggesting complementary reading recommendations (based on collaborative filtering).
  • Align inventory to demand
  1. Use statistical models based on demographic profiles of localities and historical title sales to better allocate inventory to stores.
  2. Dynamically trans-ship best-seller inventory between stores based on matching with demand patterns (since most demand for best-sellers occurs in the first week after release)
More later...

Thursday, November 27, 2008

My thoughts on Nielsen's "Top 10 Retailer Mistakes on Black Friday"

Nielsen published an article earlier this week on "Top 10 Retailer Mistakes on Black Friday".
According to Nielsen, the mistakes include:
  • Sticking to traditional categories
  • Not having a retailing objective
  • Not measuring your objectives
  • Leaving a bad first impression with new shoppers
  • Missing loyalty opportunities
  • Sticking to Friday morning
  • Not having door-buster merchandise in stock
  • No sense of urgency
  • Shallow discounts on door-busters
  • Not scouting the competition
While I agree with most of these points, I do have some disagreements:
  1. It may be a bad idea for retailers to try out new categories. I think it is OK for an electronics retailer to offer free soda and chips during Black Friday, but selling these items is a very different proposition.
  2. Most retailers do have sales objectives (by category) for Black Friday. I think what they also need to have are gross margin objectives. Given the hectic pace of Black Friday, it is often not possible to have targets for 'new customer footfalls'.
  3. Retailers need to be careful about passing discounts to the most loyal customers. Many of them are willing to shop with the retailer without the discount. On the other hand, many Black Friday shoppers are active discount seekers.
Here's wishing all of you a happy Thanksgiving

Wednesday, November 26, 2008

The brand new face(s) of Pepsi ;-)

Pepsi announced that it is rejigging its brands at the Morgan Stanley Global consumer & retail conference last week. The key elements of the brand redesign and repositioning are:

  • New logos for all the Pepsi beverage brands which supposedly leverage Pepsi's design heritage, humanity, and simplicity. The logo re-design efforts are being led by Arnell Group (the same agency that re-branded Reebok as Rbk, and Donna Karan as DKNY).
  • Clear sub-segmentation of the hydration 'need state' into simple hydration (Aquafina), body wellness (Propel, SoBe), and performance hydration (G2, Gatorade). Pepsi seeks to expand ('democratize') the brand appeal of Gatorade from athletes to exercisers.
  • Repositioning of Tropicana for the breakfast occasion and Trop50 for calorie-conscious consumers
  • Targeting of Lipton and Tazo RTD teas at specific demographic and need sub-subsegments within the Nourish need state
  • Launching hybrids of the energy drink AMP (with tea, lemon) to incorporate consumer feedback of unfamiliar taste and lack of thirst-quenching in the Transform need state
I think these are bold moves that will revitalize the CSD market and well as position Pepsi well in the non-carbonated beverages market.

Some of the products launched by Pepsi recently are very notable (Diet Pepsi Max, flavored AMP variants). As a traditional Diet Coke fan, I was surprisingly attracted to the taste of Diet Pepsi Max. I personally witnessed a free sample promotion for Diet Pepsi Max outside Grand Central station in NYC in mid-October and the reviews appeared good. Pepsi has opened up a new segment in the CSD market by adding ginseng to cola. This will attract consumers who want a rush of energy with few calories at a price point below energy drinks.

The ball is now in Coca Cola's court...

Wednesday, November 19, 2008

Black Friday deals: Do they work for the retailer?

The US holiday retail season this year is expected to be the worst in the past few decades. Many retailers have started offering Black Friday discounts in advance.

The website Black Friday Ads tracks some of the hottest discount deals being offered by retailers. They have a page which tracks the scanned advertisements of most big US retailers. Some retailers are offering online deals at prices below Black Friday prices.

I couldn't help wondering how retailers can use analytics to maximize their Black Friday sales and gross margins. Here are some suggestions :
  • Designing the black friday deals:
  1. Analyze market surveys to understand what products people want to buy during Black friday (computers, HD TVs, iPods, hard disk drives, digital cameras, GPS devices, clothes, toys) this year
  2. Benchmark prices against competition based on information available daily to ensure that you have enough traffic-pullers...these get the people into the stores.
  3. Ensure that your deals are spread across the aisles/ departments. This ensures that customers are exposed to more categories whil picking up the doorbuster items and that all doorbuster products are not picked up just by the first few customers. An analysis of previous years' transaction data will yield numbers on how many customers got a doorbuster product (typically a few hundred per store) and how each of these customers moved through the aisles.
  • After Black Friday: Analyze the baskets of transactions during Black Friday:
  1. Customers who buy doorbuster products often pick up other items that may not have significant discount. Calculate the sales and gross margin on baskets that have a doorbuster product. Retailers often lose money on the first customers in line.
  2. Customers often come in after the doorbuster products are sold (Each store usually stocks 5-15 units of each doorbuster). They often pick up products with lower discounts. To calculate the halo effect of the discount, calculate the sales and gross margin lift for Black Friday baskets without a doorbuster product
  3. So long as the total of the above has positive sales and margin impact, the retailer made money on the Black Friday deals.
  4. The learning from Black Friday can be used to better structure deals for the rest of the holiday season.

This season, retailers may have to resort to a long period of discounting with dynamic moves based on competition and sales performance on a daily basis. Getting the process of discount pricing will be very critical.

About this blog

This is my first blog post. So I wanted to write about what this blog is going to be about. Hopefully this helps you decide whether to subscribe to this blog or not.

This blog will attempt to highlight how companies can make smarter business decisions using analytics (which I define as the 'collation, summarization, mining and analysis of data').


Why this blog?


Most companies generate and collect a lot of data in their business...but don't use this data enough while making business decisions. (This may sound cliched but it is true!). My hypotheses on why this happens are:
  • Data is stored across functional and divisional silos across the company. E.g., Surveys done by the marketing team may not be available to the customer service team
  • It requires a lot of effort to clean and update data required for analysis. Corporate IT is often short of resources for maintaining analytical databases. E.g., Many B2B companies don't have a single view of a customer across all their internal and third-party databases
  • Analysis often requires ad-hoc queries of complex databases. Companies don't have enough skilled resources to query these databases
  • Some advanced analysis techniques require significant expertise (statistics, operations research, etc), computing power and time. Some managers believe it is important to take half-correct decisions quickly rather than better decisions with more time.


What industries will this blog cover?

This blog will initially focus on business problems in the retail, consumer goods, telecoms, banking and insurance industries (because I am most interested in these industries). Of course, I will include business issues from other industries if I find them interesting and if I feel I am competent enough to write about them.



What kind of business problems will this blog discuss?

I plan to write about a wide variety of business problems ranging from marketing, merchandising to operations (supply chain, logistics) to fraud detection.



Of course, I welcome comments on what topics readers would like to discuss in the blog.